Unmanned aerial vehicle target detection method based on edge intelligence

Abstract

The invention provides an unmanned aerial vehicle target detection method based on edge intelligence. The method comprises the steps of deploying a two-stage filter and a lightweight target detectionmodel S_YOLOv3 at an unmanned aerial vehicle end, and deploying a high-precision target detection model YOLOv3 at a cloud end, respectively inputting an unmanned aerial vehicle data set into S_YOLOv3and YOLOv3, and carrying out model migration training, so that the S_YOLOv3 meets a preset speed requirement, and the YOLOv3 meets a preset precision requirement, preliminarily filtering the acquiredunmanned aerial vehicle traffic video through a two-stage filter deployed at an unmanned aerial vehicle end, and discarding redundant frames, inputting the remaining frames into a lightweight vehicledetection model S_YOLOv3 to further screen target frames, and inputting the target frame into a high-precision target detection model YOLOv3 deployed at the cloud, and performing high-precision detection to obtain a final target frame. The unmanned aerial vehicle target is detected through an edge cloud cooperation mechanism, the problems of time delay, target frame loss and the like caused by overlarge data transmission quantity are solved, and the target detection speed and accuracy are improved.

Description

technical field [0001] The invention relates to the technical fields of computer vision, deep learning and mobile edge computing, and in particular to an edge intelligence-based detection method for unmanned aerial vehicles and vehicles. Background technique [0002] With the development of transportation infrastructure, many cities in China have installed thousands of traffic monitoring devices in urban areas. However, these videos are always transmitted to the monitoring center and analyzed manually, which is costly and ineffective. At present, the transportation sector uses drones to assist operations, which has become a promising technology due to its advantages of low cost, small size, flexibility and convenience. But the main function of drones is only to shoot video, which is finally analyzed by humans. Therefore, the real "unmanned" UAV is mainly faced with the following difficulties: First, the number of aerial video frames is huge, and if each frame in the video ...

AI Technical Summary

Problems solved by technology

[0003] The technical problem to be solved by the present invention is to provide an edge intelligence-based UAV vehicle target detection method, which detects the UAV target through the edge-cloud coordination mechanism, and solves the problem of time extension and lost targets caused by excessive data transmission Frame and other issues, improve the target detection speed, while improving the detection accuracy

Images